Sunday, October 09, 2016
Automating Data Science
Thoughts on automated data science. In KDNuggets. Interview with SKLearn team. Some good thoughts here. I would note that this was thought of decades ago, and it has not progressed significantly until recently, it still requires statistical and business need checking. Even partial results can be useful in getting insights about the business decision process. Also needs checking to see if the results are stable over time, data and metadata context.